1. Field of the Invention
This invention relates to display systems. More specifically, this invention relates to systems for displaying graphical information to physicians.
2. Description of the Related Technology
Medical display systems provide information to doctors in a clinical setting. Typical display systems provide data in the form of numbers and one-dimensional signal waveforms that must be assessed, in real time, by the attending physician. Alarms are sometimes included with such systems to warn the physician of an unsafe condition, e.g., a number exceeds a recommended value. In the field of anesthesiology, for example, the anesthesiologist must monitor the patient's condition and at the same time (i) recognize problems, (ii) identify the cause of the problems, and (iii) take corrective action during the administration of the anesthesia. An error in judgment can be fatal.
Approximately 50 percent of the more than 2000 anesthesia-related deaths per year have been found to be due to improper choices during surgery. In general, human error in anesthesia represents failure by the anesthesiologist to recognize a problem (abnormal physiology), identify the cause of the problem and take appropriate corrective action when administering an anesthetic to a patient. Anesthesia performance models, models showing the relationship between errors, incidents and accidents, and models depicting accident evolution in the anesthesia all illustrate the fact that anesthesia is a complex environment prone to errors.
Physiologic data displays of the patient's condition play a central role in allowing anesthesiologists to observe problem states in their patients and deduce the most likely causes of the problem state during surgery. As one might predict, 63 percent of the reported incidents in the Australian Incident Monitoring Study (AIMS) database were considered detectable with standard data monitors. Others have attempted to address these problems, but with only limited success.
For example, Cole, et. al. has developed a set of objects to display the respiratory physiology of intensive care unit (ICU) patients on ventilators. This set of displays integrates information from the patient, the ventilator, rate of breathing, volume of breathing, and percent oxygen inspired. Using information from object displays, ICU physicians made faster and more accurate interpretations of data than when they used alphanumeric displays. Cole published one study that compared how physicians performed data interpretation using tabular data vs. printed graphical data. However, Cole's work did not involve a system for receiving analog data channels and driving a real-time graphical display on a medical monitor.
In addition, Ohmeda, a company that makes anesthesia machines, manufacturers the Modulus CD machine which has an option for displaying data in a graphical way. The display has been referred to as a glyph. Physiologic data is mapped onto the shape of a hexagon. Six data channels generate the six sides of the hexagon. Although this display is graphical, the alphanumeric information of the display predominates. There is no obvious rational for why the physiologic data is assigned a side of the hexagon. Moreover, symmetric changes to the different signs of this geometric shape are very hard for individuals to differentiate.
In the surgical and postoperative settings, decisions regarding the need for blood transfusion normally are guided by hemoglobin (Hb) or hematocrit levels (Hct). Hematocrit is typically defined as the percentage by volume of packed red blood cells following centrifngation of a blood sample. If the hemoglobin level per deciliter of blood in the patient is high, the physician can infer that the patient has sufficient capacity to carry oxygen to the tissue. During an operation this value is often used as a trigger; i.e. if the value falls below a certain point, additional blood is given to the patient. While these parameters provide an indication of the arterial oxygen content of the blood, they provide no information on the total amount of oxygen transported (or "offered") to the tissues, or on the oxygen content of blood coming from the tissues.
For example, it has been shown that low postoperative hematocrit may be associated with postoperative ischemia in patients with generalized atherosclerosis. Though a number of researchers have attempted to define a critical Hct level, most authorities would agree that an empirical automatic transfusion trigger, whether based on Hb or Hct, should be avoided and that red cell transfusions should be tailored to the individual patient. The transfusion trigger, therefore, should be activated by the patient's own response to anemia rather than any predetermined value.
This is, in part, due to the fact that a number of parameters are important in determining how well the patient's tissues are actually oxygenated. In this regard, the patient's cardiac output is also an important factor in correlating hemoglobin levels with tissue oxygenation states. Cardiac output or CO is defined as the volume of blood ejected by the left ventricle of the heart into the aorta per unit of time (ml/min) and can be measured with thermodilution techniques. For example, if a patient has internal bleeding, the concentration of hemoglobin in the blood might be normal, but the total volume of blood will be low. Accordingly, simply measuring the amount of hemoglobin in the blood without measuring other parameters such as cardiac output is not always sufficient for estimating the actual oxygenation state of the patient.
More specifically the oxygenation status of the tissues is reflected by the oxygen supply/demand relationship of those tissues i.e., the relationship of total oxygen transport (DO.sub.2) to total oxygen consumption (VO.sub.2). Hemoglobin is oxygenated to oxyhemoglobin in the pulmonary capillaries and then carried by the cardiac output to the tissues, where the oxygen is consumed. As oxyhemoglobin releases oxygen to the tissues, the partial pressure of oxygen (PO.sub.2) decreases until sufficient oxygen has been released to meet the oxygen consumption (VO.sub.2). Although there have been advances in methods of determining the oxygenation status of certain organ beds (e.g., gut tonometry; near infrared spectroscopy) these methods are difficult to apply in the clinical setting. Therefore, the use of parameters that reflect the oxygenation status of the blood coming ftQm the tissues i.e., the partial pressure of oxygen in the mixed venous blood (PvO.sub.2 ; also known as the mixed venous blood oxygen tension) or mixed venous blood oxyhemoglobin saturation (SvO.sub.2) has become a generally accepted practice for evaluating the global oxygenation status of the tissues.
Unfortunately, relatively invasive techniques are necessary to provide more accurate tissue oxygenation levels. In this respect, direct measurement of the oxygenation state of a patient's mixed venous blood during surgery may be made using pulmonary artery catheterization. To fully assess whole body oxygen transport and delivery, one catheter (a flow directed pulmonary artery [PA] catheter) is placed in the patient's pulmonary artery and another in a peripheral artery. Blood samples are then drawn from each catheter to determine the pulmonary artery and arterial blood oxygen levels. As previously discussed, cardiac output may also be determined using the PA catheter. The physician then infers how well the patient's tissue is oxygenated directly from the measured oxygen content of the blood samples.
While these procedures have proven to be relatively accurate, they are also extremely invasive. For example, use of devices such as the Swan-Ganzg.RTM. thermodilution catheter (Baxter International, Santa Ana, Calif.) can lead to an increased risk of infection, pulmonary artery bleeding, pneumothorax and other complications. Further, because of the risk and cost associated with PA catheters, their use in surgical patients is restricted to high-risk or high-blood-loss procedures (e.g., cardiac surgery, liver transplant, radical surgery for malignancies) and high-risk patients (e.g., patients who are elderly, diabetic, or have atherosclerotic disease).
Among other variables, determination of the oxygenation status of the tissues should include assessment of the amount of blood being pumped toward the tissues (CO) and the oxygen content of that (arterial) blood (CaO.sub.2). The product of these variables may then be used to provide a measure of total oxygen transport (DO.sub.2). Currently, assessment of DO.sub.2 requires the use of the invasive monitoring equipment described above. Accordingly, determination of DO.sub.2 is not possible in the majority of surgical cases. However, in the intensive care unit (ICU), invasive monitoring tends to be a part of the routine management of patients; thus, DO.sub.2 determinations are obtained more readily in this population.
Partial pressure of oxygen in the mixed venous blood or mixed venous blood oxygen tension (PvO.sub.2) is another important parameter that may be determined using a PA catheter. Because of the equilibrium that exists between the partial pressure of oxygen (PO.sub.2) in the venous blood and tissue, a physician can infer the tissue oxygenation state of the patient. More specifically, as arterial blood passes through the tissues, a partial pressure gradient exists between the PO.sub.2 of the blood in the arteriole passing through the tissue and the tissue itself. Due to this oxygen pressure gradient, oxygen is released from hemoglobin in the red blood cells and also from solution in the plasma; the released O.sub.2 then diffuses into the tissue. The PO.sub.2 of the blood issuing from the venous end of the capillary cylinder (PvO.sub.2) will generally be a close reflection of the PO.sub.2 at the distal (venous) end of the tissue through which the capillary passes.
Closely related to the mixed venous blood oxygen tension (PvO.sub.2) is the mixed venous blood oxyhemoglobin saturation (SvO.sub.2) which is expressed as the percentage of the available hemoglobin bound to oxygen. Typically, oxyhemoglobin disassociation curves are plotted using SO.sub.2 values vs. PO.sub.2 values. As the partial pressure of oxygen (PO.sub.2) decreases in the blood (i.e. as it goes through a capillary) there is a corresponding decrease in the oxygen saturation of hemoglobin (SO.sub.2). While arterial values of PO.sub.2 and SO.sub.2 are in the neighborhood of 95 mm Hg and 97% respectively, mixed venous oxygen values (PvO.sub.2, SvO.sub.2) are on the order of 45 mm Hg and 75% respectively. As such SvO.sub.2, like PvO.sub.2, is indicative of the global tissue oxygenation status. Unfortunately, like PvO.sub.2, it is only measurable using relatively invasive measures.
Another rather informative parameter with respect to patient oxygenation is deliverable oxygen (dDO.sub.2 ). dDO.sub.2 is the amount of the oxygen transported to the tissues (DO.sub.2 ) that is able to be delivered to the tissues (i.e. consumed by the tissues) before the PvO.sub.2 (and by implication the global tissue oxygen tension) falls below a certain value. For instance the dDO.sub.2 (40) is the amount of oxygen that can be delivered to the tissues (consumed by the tissues) before PvO.sub.2 is 40 mm Hg while dDO.sub.2 (35) is the amount consumed before the PvO.sub.2 falls to 35 mm Hg.)
Additional relevant parameters may be determined non-invasively. For instance, whole body oxygen consumption (VO.sub.2) can be calculated from the difference between inspired and mixed expired oxygen and the minute volume of ventilation. Cardiac output may also be non-invasively inferred by measuring arterial blood pressure instead of relying on thermodilution catheters. For example, Kraiden et al. (U.S. Pat. No. 5,183,051, incorporated herein by reference) use a blood pressure monitor to continuously measure arterial blood pressure. These data are then converted into a pulse contour curve waveform. From this waveform, Kraiden et al. calculate the patient's cardiac output.
Regardless of how individual parameters are obtained, those skilled in the art will appreciate that various well established relationships allow additional parameters to be derived. For instance, the Fick equation (Fick, A. Wurzburg, Physikaisch edizinische Gesellschaft Sitzungsbericht 16 (1870)) relates the arterial oxygen concentration, venous oxygen concentration and cardiac output to the total oxygen consumption of a patient and can be written as: EQU (CaO.sub.2 -CvO.sub.2).times.CO=VO.sub.2
where CaO.sub.2 is the arterial oxygen content, CvO.sub.2 is the venous oxygen content, CO is the cardiac output and VO.sub.2 represents whole body oxygen consumption.
While the non-invasive derivation of such parameters is helpful in the clinical setting, a more determinative "transfusion trigger" would clearly be beneficial. If PvO.sub.2 or DO.sub.2 is accepted as a reasonable indicator of patient safety, the question of what constitutes a "safe" level of these parameters arises. Though data exists on critical oxygen delivery levels in animal models, there is little to indicate what a critical PvO.sub.2 might be in the clinical situation. The available data indicate that the level is extremely variable. For instance, in patients about to undergo cardiopulmonary bypass, critical PvO.sub.2 varied between about 30 mm Hg and 45 mm Hg where the latter value is well within the range of values found in normal, fit patients. Safe DO.sub.2 values exhibit similar variability.
For practical purposes a PvO.sub.2 value of 35 mm Hg or more may be considered to indicate that overall tissue oxygen supply is adequate, but this is implicit on the assumption of an intact and functioning vasomotor system. Similarly, the accurate determination of DO.sub.2 depends on an intact circulatory system. During surgery it is necessary to maintain a wide margin of safety and probably best to pick a transfusion trigger (whether DO.sub.2 , PvO.sub.2 , SvO.sub.2 or some derivation thereof) at which the patient is obviously in good condition as far as oxygen dynamics are concerned. In practice, only certain patients will be monitored with a pulmonary artery catheter. Accordingly, the above parameters will not be available for all patients leaving the majority to be monitored with the imperfect, and often dangerous, trigger of Hb concentration.
Efforts to resolve these problems in the past have not proven entirely successful. For example, Faithfill et al. (Oxygen Transport to Tissue XVI, Ed. M. Hogan, Plenum Press, 1994, pp. 41-49) describe a model to derive the oxygenation status of tissue under various conditions. However, the model is merely a static simulation allowing an operator to gauge what effect changing various cardiovascular or physical parameters will have on tissue oxygenation. No provisions are made for continuous data acquisition and evaluation to provide a dynamic representation of what may actually be occurring. Accordingly, the model cannot be used to provide real-time measurements of a patient's tissue oxygenation under changing clinical conditions.
Thus, what is needed in the art is a relatively non-invasive system for intuitively displaying physiological information to a physician. The system described below provides such a system to improve a physician's interpretation of patient data. Other aspects of the system will become apparent in the description that follows.